Adaptive multioutput gradient RBF tracker for nonlinear and nonstationary regression
Adaptive multioutput gradient RBF tracker for nonlinear and nonstationary regression
Multioutput regression of nonlinear and nonstationary data is largely understudied in both machine learning and control communities. This article develops an adaptive multioutput gradient radial basis function (MGRBF) tracker for online modeling of multioutput nonlinear and nonstationary processes. Specifically, a compact MGRBF network is first constructed with a new two-step training procedure to produce excellent predictive capacity. To improve its tracking ability in fast time-varying scenarios, an adaptive MGRBF (AMGRBF) tracker is proposed, which updates the MGRBF network structure online by replacing the worst performing node with a new node that automatically encodes the newly emerging system state and acts as a perfect local multioutput predictor for the current system state. Extensive experimental results confirm that the proposed AMGRBF tracker significantly outperforms existing state-of-the-art online multioutput regression methods as well as deep-learning-based models, in terms of adaptive modeling accuracy and online computational complexity.
Adaptation models, Adaptive systems, Computational modeling, Data models, Mathematical models, Multioutput gradient radial basis function (MGRBF) network, multivariate nonlinear and nonstationary regression, online adaptive tracking, Predictive models, Training, two-step training
7906-7919
Liu, Tong
0b4a852b-76d0-4555-8f6c-2ccf76ab4fbf
Chen, Sheng
9310a111-f79a-48b8-98c7-383ca93cbb80
Li, Kang
4531da99-f15c-4e7a-9991-c2b5c63019c6
Gan, Shaojun
458954ae-151d-4355-a677-3470632569f7
Harris, Chris J.
daa59e88-2e26-42df-bf2e-f8e6792ecb18
1 December 2023
Liu, Tong
0b4a852b-76d0-4555-8f6c-2ccf76ab4fbf
Chen, Sheng
9310a111-f79a-48b8-98c7-383ca93cbb80
Li, Kang
4531da99-f15c-4e7a-9991-c2b5c63019c6
Gan, Shaojun
458954ae-151d-4355-a677-3470632569f7
Harris, Chris J.
daa59e88-2e26-42df-bf2e-f8e6792ecb18
Liu, Tong, Chen, Sheng, Li, Kang, Gan, Shaojun and Harris, Chris J.
(2023)
Adaptive multioutput gradient RBF tracker for nonlinear and nonstationary regression.
IEEE Transactions on Cybernetics, 53 (12), .
(doi:10.1109/TCYB.2023.3235155).
Abstract
Multioutput regression of nonlinear and nonstationary data is largely understudied in both machine learning and control communities. This article develops an adaptive multioutput gradient radial basis function (MGRBF) tracker for online modeling of multioutput nonlinear and nonstationary processes. Specifically, a compact MGRBF network is first constructed with a new two-step training procedure to produce excellent predictive capacity. To improve its tracking ability in fast time-varying scenarios, an adaptive MGRBF (AMGRBF) tracker is proposed, which updates the MGRBF network structure online by replacing the worst performing node with a new node that automatically encodes the newly emerging system state and acts as a perfect local multioutput predictor for the current system state. Extensive experimental results confirm that the proposed AMGRBF tracker significantly outperforms existing state-of-the-art online multioutput regression methods as well as deep-learning-based models, in terms of adaptive modeling accuracy and online computational complexity.
Text
MGRBF_R1
- Accepted Manuscript
Text
Adaptive_Multioutput_Gradient_RBF_Tracker_for_Nonlinear_and_Nonstationary_Regression
- Version of Record
Restricted to Repository staff only
Request a copy
More information
Accepted/In Press date: 3 January 2023
e-pub ahead of print date: 2 February 2023
Published date: 1 December 2023
Additional Information:
Funding Information:
This work was supported by the National Natural Science Foundation of China under Grant 62003011.
Publisher Copyright:
© 2013 IEEE.
Keywords:
Adaptation models, Adaptive systems, Computational modeling, Data models, Mathematical models, Multioutput gradient radial basis function (MGRBF) network, multivariate nonlinear and nonstationary regression, online adaptive tracking, Predictive models, Training, two-step training
Identifiers
Local EPrints ID: 474173
URI: http://eprints.soton.ac.uk/id/eprint/474173
ISSN: 2168-2267
PURE UUID: dab1d2d9-920d-415c-ba42-d9bf858c2ace
Catalogue record
Date deposited: 14 Feb 2023 18:00
Last modified: 17 Mar 2024 00:02
Export record
Altmetrics
Contributors
Author:
Tong Liu
Author:
Sheng Chen
Author:
Kang Li
Author:
Shaojun Gan
Author:
Chris J. Harris
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics